Percify Studio

No. Percify does not apply a watermark to output on any plan, including the free one, and you own what you generate on every plan. Paid plans differ by how many credits they carry, not by watermarking or licensing.

Because text-to-video has nothing to anchor to. The model invents the subject, the motion and the scene together, so the same prompt run twice returns different people and places. If the subject has to stay the same between shots, start from an image instead: image-to-video keeps the person in the photo fixed and generates motion around them.

Usually because the prompt describes a style rather than a scene. Prompts like “cinematic” or “professional” give the model no subject, framing or motion to work from, and it fills the gaps itself. Name the shot instead: who is in frame, how they are framed, what moves, and how it is lit. The more of the scene you specify, the less the model invents.

Generated motion accumulates error over time, so the opening seconds hold together better than the end. Faces and hands show it first because small distortions are most visible there. Generate in shorter segments and join them rather than asking for one long take, and frame tighter so there is less for the model to invent.

Percify does not have per-seat accounts today; a plan is one login, so a team shares one account and its credits.

Videos download as MP4. Resolution depends on the model: talking videos render at up to 720p, and each text and image to video model shows its own resolutions before you run it.